Abstract: The research uses the big data of mobile phone signaling to locate the visitors of large city park in Shanghai during the active outdoor recreation period. The locations of park visitors' homes have been found. And it is able to describe park's "real" service area. The visitors sample data in the "real" service areas of the 32 large urban parks are connected with the Shanghai street space unit, and the park emissiveness of the units are calculated according to the numbers of park visitors whose home location is in the unit and the population in the corresponding unit. With the recreation demand expressed by emissiveness combined with the park supply expressed by park location, Shanghai can be divided into four types of areas which are high demand for low supply, high demand for high supply, low demand for high supply, low demand for low supply. Each type of park service characteristics have been described. According to the supplies of large and community parks in different areas, this paper puts forward some suggestions on the location and construction priorities of the new large parks. And the order of construction priority of regional park layout in "Shanghai Master Plan 2016-2040" is also suggested.